New Topographies - Images of an Algorithmic Landscape

A post-photographic survey of the topographies of latent space.

Peering out from a vantage point at the edgelands of Baudrillard's 'desert of the real', familiar yet foreign lands take form and come into view. Journeying into this metaphysical world as part of a contemporary topographical study, all the images in this experimental survey were created using an artificial intelligence (AI) text-to-image generator. The AI was 'prompted' by inputting the titles of the photographs featured in the original 1975 catalogue from the seminal exhibition 'New Topographics: Photographs of a Man-Altered Landscape'. Curated by William Jenkins and featuring photographs by Robert Adams, Lewis Baltz, Bernd and Hilla Becher, Joe Deal, Frank Gohlke, Nicholas Nixon, John Schott, Stephen Shore and Henry Wessel, Jr., the groundbreaking exhibition surveyed the vernacular landscape and the highly influential documentary approaches employed in its photographic rendering.

Presently, most commercially available AI image generators are trained on vast data sets of images, most of which include broad 'scrapes' of the internet. Given the sheer scale of data mining as part of the ongoing machine learning enterprise, these scrapes possibly include elements of the online digital archives and/or digitised reproductions of the photographers' works featured in the original exhibition, as well as the myriad topographical photographs they inspired and continue to inspire (though as a typical end-user of a commercial AI image generator there is no direct way of knowing the exact content of the data set it was trained on).

Closely resembling the form and structure of the original exhibition catalogue, New Topographies: Images of an Algorithmic Landscape presents a small but visually diverse sample of the imagery that was produced during this initial exploration. Like most emerging technologies, the view of the horizon that AI imaging currently affords is not yet fully discernible. However, as the algorithms' rendering capabilities invariably improve (which is happening at pace), and as they become more mainstream in their adoption, many ethical concerns are likely to become apparent, with divisive issues around authorship and copyright, creative enterprise and artistic merit, environmental sustainability, as well as the impact of AI produced deepfake imagery already starting to loom large. In the meantime, a growing number of users are utilising these increasingly powerful imaging tools and beginning the process of traversing and mapping out this new, unchartered terrain.